When it comes to antennas and antenna "gain", physics are the limiting factor that can't be overcome by marketing claims and "my antenna is better than yours" claims. People who do radio professionally know this. And although this scanner thing is just a hobby, if what you are asking for is the best reception you can get (within your budget) given the limitations of antenna placement, you need to put together something called a "link budget". This is where you breakdown each component of the antenna system and calculate each gain or loss on the frequencies in question. By properly considering all these pieces of the "link", you can pretty much know what to expect at the end that connects to the radio.
It starts at the antenna. This is the only place you can expect natural "gain". Three factors need to be considered here.
1) Resonance on the frequencie(s) desired. If it has a high return loss (bad VSWR) at any frequency you may want to hear, don't expect good results. Antennas that use coils as part of their design often are narrow band and have dead spots that really degrade frequencies outside the narrow resonance areas they are "tuned" to.
2) Angle of radiation on said frequency or frequencies. This is often where gain is realized by focusing more energy in the angle and direction where you need to RX and maybe TX. This comes at the expense of sacrificing other angles or directions, and this is where size matters and can't be physically overcome by gimmicky marketing claims or "my antenna is the best" claims without reviewing the full specifications at all intended frequencies. For omnidirectional antennas, the only way to get true gain is by stacking multiple elements of dipoles (open, folded, or collinear in design). But then bandwidth is a major limitation and can only work well on frequencies within physical limitations (single band elements). When considering multi-band antennas, be careful to look at all the data concerning frequencies covered without nulls, angles of radiation at all claimed frequencies covered (viewing polar charts of the antenna in question...if the company even published that vital information) , and deceptive gain figures (marketing lies). A small antenna claiming big gain on low frequencies (or worse yet all "bands") is a red flag. If real commercial antennas (the ones we are listening to at tower sites) are 21' tall and have 6 DB of gain, your little 4' antenna can't have anywhere near that gain. It's the law of physics!
3) Height is might. It's mostly about line of site, so height matters to clear obstructions and work out to the curvature of the earth. WARNING: Preamps used as a gain substitute for actual natural physical gain mostly cause major problems with scanners, as they overload the wideband design of the reciver and produce false images. And those are also subject to specifications...like noise figures and overload limitations. The bottom line is you can't expect good results without considering ALL of the physical elements and user requirements. That is the "real deal", and everything else is rubbish.
No matter what you use for feedline, it's a loss. Now the real question is, "how much loss can I tolerate?" Well, considering you just (hopefully) chose the ideal antenna, do you really want to throw away the gains you just made? Probably not! So, you are wise to not use cheap junk feedline as an easy way out, or you'll do just that...throw away perfectly good signals you could have easily received. Feedline is the one area we can usually get hard data of DB loss per foot at given frequency ranges. You can start calculating your losses and might be horrified at the results, especially at the all-so-popular 800 MHz range. This is why pros use hardline and avoid LMR. My LMR use resulted in more noise, which hurt weaker signals, than I ever expected. I use it in only very limited capacity where I can't use hardline...but don't like it. The best way to manage hardline is to terminate it to a weatherproof grounding box just where you want to enter the building. Use something like an Alpha Delta wide-band surge arrestor and ground it well. You get a place to connect smaller feedline for the final run to the scanner, and the all-so-important lightning protection! Then use a high quality, low loss, thin cable like superflex or (my favorite) silver plated / double shielded RG400. Anything like RG58, RG8X or RG6 (not even 50 Ohm cable) will cause a lot more loss than you might think. I've measured it, and was shocked at RG8X. Lost half of my power at UHF even with good connectors installed perfectly. You can't argue with physics and real test data.
Lastly, quality connectors are a must, and need to be installed perfectly. Type N connectors are very low loss, and the professional choice when DIN connectors are too large to consider. That final BNC connection to the radio is unavoidable. Not a bad connector design considering all factors. SMA, as well, is a good small cable connector. Luckily, you can readily purchase most cable and connector types professionally assembled, so you may not need the skills to do your own connector assembly. But it's a skill worth having.
So why, after all this background, do I recommend a Diamond discone? Unless you are ready to combine large single-band omnidirectional gain antennas together into a single feedline, you can't overcome the wideband low-loss design of a QULITY discone antenna. That is why the military uses them extensively. But not all (Voyager's "wet noodle" for example) are a quality design. I have compared discone antennas, and some (like the Comet) are pretty useless. Compared to real commercial gain antennas, they do better that I ever expected for a so-called "unity gain" antenna. The bottom line is I hear a very long distance consistently with modest height. The real gain antennas do a bit better, but only on frequencies they effectively cover. Discone antennas cover just about every frequency my scanner or ham gear need to cover...all in one antenna. By using a 62 3/4" whip on top, I have a wide band 6m and low band extension that I have used to talk to northern New Hampshire and Canada (multiple times...good reports) from my station here in Florida. That is 1400 miles, and if you know the limits of E Skip, that is about as low an angle...and far a distance as you can get on one hop. This proves the focus is low on the horizon and not 35 degrees up in the air. But the best part is this whip also works as a 3/4 wave on VHF high band, so my VHF performance also gets a boost. This antenna blows me away, and not just because I "think" it is the best, or make unsubstantiated claims. The science behind it says it is the practical choice for wide band omnidirectional applications. Made of quality stainless steel, my Diamond discones have lasted over 30 years without any issue. You can't go wrong, especially if you limit all the loss you can in the feedline. Do it right, and you will be rewarded!
Phil